The Analyze Data node turns your raw marketing data into summaries, insights, or charts using AI. It’s built to handle structured datasets like ad reports, campaign metrics, or website performance—so you can automate analysis, not just reporting.

​
When to Use It

Use this node to:

  • Summarize campaign or ad performance data
  • Highlight spikes, drops, or patterns in KPIs
  • Calculate new fields (like conversion rate)
  • Generate Markdown summaries or structured output
  • Power your Google Slides, Email, or Sheet-based workflows

​
Inputs

FieldTypeRequiredDescription
DataDynamic Text Area✅Insert structured data like a table or JSON (e.g. from a “Get Report” node)
PromptDynamic Text Area✅Describe what you want AI to do (e.g. “Summarize trends and calculate conversion rate”)
ModelModel Selector✅Choose a model like GPT-4.1 or GPT-4o based on your accuracy needs and credit budget
Force Python ExecutionToggle❌Forces AI to use Python to calculate fields like CTR, averages, or conversions (no hallucinated math)
Schema FieldsSchema Builder❌Define the structure of the output if you want consistent, reusable results (e.g., a table or structured list)

​
What Is Schema / Structured Output?

Normally, AI responds with a paragraph or list—but that’s not always useful when you want to reuse the output in Slides, Sheets, or another workflow.

By defining a schema, you’re telling the AI:

“Stick to this format. Return clean fields I can reuse later.”

​
Why It Matters

Structured output gives you:

  • Consistent fields, every time
  • Machine-readable format (JSON/table)
  • Reusable output for other steps (email, spreadsheet, visualizations)

​
Example: Calculate Conversion Rate with Schema

Say your previous step outputs a dataset like:

ClicksConversions
100045
120060
80032

You want AI to calculate conversion_rate (Conversions ÷ Clicks) and return a table with three columns.

Schema Configuration:

  • clicks — Number — “Number of ad clicks”
  • conversions — Number — “Number of conversions”
  • conversion_rate — Percentage — “Conversions divided by clicks (e.g. 0.045 = 4.5%)”

Now the output will be a clean table that you can pass to Google Sheets, display in a slide, or summarize in another step.

​
Force Python Execution (When You Want Guaranteed Math)

AI models sometimes get math wrong—especially with ratios, percentages, or aggregations. By default, the AI model will decide to use Python or not based on the prompt.

If you want to ensure accurate calculations, enable Force Python Execution. This forces the AI to use backend Python logic for any math operations, avoiding “hallucinations” or incorrect calculations.

Enable Force Python Execution if:

  • You want guaranteed, accurate calculations
  • You’re asking the model to aggregate or compute fields (e.g. total spend, CTR, ROI)
  • You’re building dashboards or decision-making reports

This ensures AI runs backend Python for logic instead of “guessing” in natural language.

​
Output

  • Without schema: You’ll get a markdown summary with insights, comparisons, or trends.
  • With schema: You’ll get a clean, structured JSON table you can pass to downstream nodes (Slides, Sheets, etc.)

​
Example Use Case

  1. Get Report node pulls data from Google Ads
  2. Pass that data to Analyze Data
  3. Prompt:
    “Summarize the data. Calculate conversion rate for each row. Highlight anything with below-average performance.”
  4. Turn on Force Python Execution
  5. Define a schema with: clicks, conversions, conversion_rate

Now the output will be:

  • Accurate
  • Consistently structured
  • Easy to reuse or send

​
Credit Cost

Cost depends on the selected model. See the Credits & Usage page for details.

​
FAQs